ML Series #3: Why SkLearn Makes ML 10x Easier (+ Linear Regression Pitfalls)

ML Series #3: Why SkLearn Makes ML 10x Easier (+ Linear Regression Pitfalls) In this video, we utilize sklearn to create the perceptron with fewer lines of code. However, the perceptron has limitations. To understand the limitations, we dive into its shortcoming: its inability to handle non-linear data distributions, as it relies exclusively on linear decision boundaries. To account for this, we discuss scenarios where linear classification fails and introduce logistic regression as an alternative. Logistic regression’s use of logarithmic operations enables curved decision boundaries, improving classification for more complex datasets. 0:00 Intro 1:06 Coding the Perceptron 1:45 Start Data 2:16 Import Data 3:21 Look At Data Set 5:32 Print Labels 5:52 Training And Testing Data 7:20 Standard Scaler 11:06 Percept 12:05 Set Prediction 13:44 Graph 16:01 Cons of the Perceptron 18:15 Outro ---------------------------------------------------------------------------------- Social Link: Github https://github.com/ethicalPap LinkedIn   / vankperry   Research Profile: https://orcid.org/0009-0001-5052-6882 Join our community! Discord   / discord   ---------------------------------------------------------------------------------- Business Email: ethicalpap@gmail.com ---------------------------------------------------------------------------------- Video Editor: filmzjasper@gmail.com